Retinoblastoma organoids: Reconstructing a tumour in vitro

ArticleDetailDownload PDF

Retinoblastoma is a childhood eye cancer that can cause vision loss and even death. Understanding how the tumour is initiated and how it progresses is therefore essential. Dr Zi-Bing Jin, from the Beijing Institute of Ophthalmology, developed retinoblastoma organoids, a high-fidelity three-dimensional model of the human eye cancer. It is the first time in the world that researchers successfully generate retinoblastoma in a dish. This model can be used to study the mechanisms of tumorigenesis, identify the cells that initiate the tumour, and even screen potential therapeutic drugs.

Retinoblastoma is the most prevalent ocular cancer that mostly affects children under the age of 5. It starts in the retina, the very back part of the eye that senses light, and can cause blindness. In about two out of three cases, only one eye is affected, but it can also affect both eyes.

This type of cancer does not affect everyone equally. In high-income countries, early detection and treatment allow up to 95% of affected children to survive. However, in the rest of the world, the survival rate can be lower than 30%. Blindness or vision impairment remains a common complication even in children who are cured of the disease.

Because the prognosis associated with retinoblastoma is dismal despite the best currently available therapies, it is essential to understand how this cancer develops in order to find new ways to prevent and treat it.

A genetic disease

Retinoblastoma is a genetic disease. The condition occurs when the gene named RB1 is mutated. RB1 is what is called a tumour-suppressor gene. It regulates cell growth and division, and plays an important role in preventing the development of cancer cells. Mutations cause dysfunction or even inactivation of these genes. As a result, cells grow and multiply abnormally, and accumulate to form a tumour.

“Retinal organoids provide an extraordinary research path to
model retinal diseases and test the effects of drugs.”

When mutations of RB1 are inherited from parents or occur during early development in the womb, all cells in the body have the mutation. Therefore, retinoblastoma usually affects both eyes. For these children, having this mutation in all the cells of the body represents an increased risk of developing other cancers. In about two out of three children with retinoblastoma, mutation of RB1 occurs later in a single cell that multiplies and forms a tumour in one eye.

Retinoblastoma is a life-threatening intraocular cancer with symptom of leukoconia. Fadhillah Randy/

Limits of current strategies

Over the past decades, scientists have developed various strategies to study retinoblastoma, but each of them has its limits. Mouse models have allowed researchers to uncover different characteristics of retinoblastoma. However, retinoblastoma in humans is not completely identical to retinoblastoma in mice and, therefore, mouse models lack human features.

To overcome this limit, researchers have transplanted grafts derived from human patients into mice in order to create in vivo mouse models with retinoblastoma holding human features. Still, transplantation from one species to another is challenging.

In vitro models have also proven to be useful: human immortalised cancer cell lines have been widely used to study retinoblastoma and test potential therapies. However, these models are two-dimensional and fail to represent the actual complexity of retinoblastoma and their three-dimensional environment.

Cyclin-dependent kinase 4 is responsible for the phosphorylation of RB1. ibreakstock/

Retinoblastoma organoids

Dr Zi-Bing Jin and his team from the Beijing Institute of Ophthalmology for the first time developed retinoblastoma organoids using state-of-art human three-dimensional retinal organoids. Retinal organoids are self-generated from human embryonic stem cells (cells derived from embryos which can develop into any type of cell). In vitro, stem cells multiply and differentiate into different types of cells to create a three-dimensional model of human retina. Retinal organoids provide an extraordinary research path to model retinal diseases and test the effects of drugs.

Dr Jin and his team genetically manipulated human embryonic stem cells to inactivate the gene RB1. These genetically modified stem cells then differentiated into human three-dimensional retinoblastoma organoids, something which had never been done before. The team then used these organoids to conduct different experiments.

A good model

Retinoblastoma organoids allowed the researchers to observe tumorigenesis – the development of a tumour. By comparing retinal and retinoblastoma organoids, they noted differences after the stem cells differentiated into retinal cells: in retinoblastoma organoids, tumour-like structures with an uneven interior density, ill-defined edges and a larger size were visible. These structures are similar to primary retinoblastoma tumours as observed directly in patients. This observation indicates that the organoids developed by Dr Jin’s team from mutated stem cell lines are a good model for retinoblastoma.

Photoreceptors in the retina include rod cells and cone cells. gritsalak karalak/

Molecular characteristics

In our cells, even though our DNA contains all of our genes, not all genes are expressed; DNA can be compared to a cookbook in which each gene is a recipe for a specific protein. Each cell, depending on its identity and its function, only uses a part of these recipes. This is regulated by plenty of signals that indicate which genes to express.

“Dr Jin and his team genetically manipulated human embryonic stem cells to inactivate the gene RB1.”

Compared to normal cells, gene expression in cancer cells is different: some genes are more or less expressed. For example, oncogenes (genes that have the potential to cause cancer) are upregulated, while tumour-suppressor genes (like RB1) are downregulated. This leads to molecular characteristics referred to as the genetic and epigenetic signatures of cancer cells.

To control that the retinoblastoma organoids they developed were an accurate model, the team from Beijing Institute of Ophthalmology analysed the molecular characteristics. They found out that the genetic and epigenetic signatures of organoid retinoblastoma were very similar to those of retinoblastoma as observed directly in patients. This shows that, in addition to having similar structures, retinoblastoma organoids are also an accurate model of the disease at the molecular level.

Retinoblastoma tumorigenesis: Green indicates the Ki67 marker, which is the proliferation marker; red indicates the SYK marker, which is an oncogene marker and can also be used as a Rb marker; blue indicates the nucleus.

Tumorigenicity of cancerous cells

Maintaining cells alive is not always easy because all the conditions required for them to survive must be reunited. After ten weeks in vitro, retinoblastoma organoids retained viability and expanded rapidly. However, just because cells thrive in vitro does not mean they have the ability to survive and expand in vivo, as more environmental constraints may apply there.

To test the viability of retinoblastoma organoids in vivo, Dr Jin and his team engrafted the cells into the eyes of mice. They observed that, two months after the injection, tumours developed. These tumours had the structural characteristics and genetic signatures of retinoblastoma. These results demonstrate that retinoblastoma organoids could retain viability and expand in vivo.

Cell of origin

The retina is made of nine different types of cells, including cone and rod cells which receive light. These nine different types of cells can be found in healthy retinal organoids. In retinoblastoma organoids, however, Dr Jin and the team noticed additional types of cells such as retinoblastoma cells and retinoma-like cells (retinoma is the benign precursor of retinoblastoma). They also noticed that the number of normal retinal cells was dramatically reduced while the number of cone precursors was excessively high. The researchers suggested that retinoma-like cells could be an intermediate cell stage between malignant cone precursors and retinoblastoma cells.

Among these different cells, they wanted to identify those that initiated the tumour. To do this, they analysed the molecular characteristics of the organoid cells over time to monitor the changes occurring in the different cell types. They observed a change in maturing cone precursors (cells destined to become cone cells): the molecular characteristics of these cells indicated that, while these cells normally do not proliferate, they are proliferative in retinoblastoma organoids.

These results led Dr Jin’s team to conclude that the tumour in retinoblastoma organoids originates from maturing cone precursor cells. They suggest that the inactivation of the RB1 gene induces an unusual activation of the cell cycle in maturing cone precursors. This activation changes them into retinoma-like cells which undergo undue proliferation to eventually transform into retinoblastoma.

Testing potential therapeutic drugs

After proving that retinoblastoma organoids were a high-fidelity model of retinoblastoma and using this model to identify the cells that initiate the tumour, Dr Jin and the team tested four chemotherapeutic drugs currently in clinical use. Most of these drugs caused a significant reduction in the number of cancer cells in retinoblastoma organoids. This result suggests that retinoblastoma organoids are an efficient tool that could be used in the future to screen potential therapeutic drugs.

In brief, Dr Jin and colleagues have established a novel human organoid retinoblastoma model derived from human embryonic stem cells in-a-dish, delineated the cancerous cell-of-origin and tested potential therapeutic agents. Their research provides new insights into cancerous origin and therapeutic targets of other human cancers.

What led you to be interested in retinoblastoma?

Retinoblastoma is the most common intraocular malignant tumour and there is no effective disease model. Therefore, we suppose to use our retinal organoids differentiation system to generate the tumour in a dish, in order to benefit the study as well as the drug screening of retinoblastoma.



  • Liu, H., Zhang, Y., Zhang, Y. Y., Li, Y. P., Hua, Z. Q., Zhang, C. J., Wu, K. C., Yu, F., Zhang, Y., Su, J., & Jin, Z. B. (2020). Human embryonic stem cell-derived organoid retinoblastoma reveals a cancerous origin. Proceedings of the National Academy of Sciences of the United States of America, 117(52), 33628–33638. Available at:

Research Objectives

Dr Jin has tested novel candidate therapeutic agents for human retinoblastoma.


Institute of Biomedical Big Data, School of Biomedical Engineering, School of Ophthalmology and Optometry, The Eye Hospital, Wenzhou Medical University


Zi-Bing Jin is the director of Beijing Institute of Ophthalmology, full professor of Capital Medical University (CMU) and chief physician at Beijing Tongren Hospital, CMU. Dr Jin focuses on stem cell translational medicine in retinal health and genetic mechanisms of ocular diseases. His team is dedicated to elucidating the disease mechanisms of inherited retinal degeneration and children ocular disorders, translating laboratory technology to improve bedside outcome, and solving key problems around this disorder.

Zi-Bing Jin

Zi-Bing Jin
Beijing Institute of Ophthalmology, Beijing Tongren Eye Center, Beijing Tongren Hospital, Capital Medical University, Beijing Ophthalmology and Visual Sciences Key Laboratory, 100730 Beijing, China


Related posts.


Leave a Comment

Your email address will not be published. Required fields are marked *

Share this article.

Share on facebook
Share on twitter
Share on whatsapp
Share on linkedin
Share on reddit
Share on email